On Mon, 2 Jun 2008, Sandy Thatcher wrote:
It's not urgently needed (as research impact is) but it is definitely a bonus too. You asked what the institution gains, and I gave a list of 10 things, some primary, some secondary.SH: OA IRs provide free supplemental copies of published, refereed journal articles. The best and most likely way to find and access them is via a harvester/indexer that links to the item, not by directly searching the IR itself. (Direct searching of the IR is more relevant for (1) institution-internal record-keeping, (2) performance assessment, (3) CV-generation, (4) grant application/fulfillment, and perhaps also some window-shopping by prospective (5) faculty, (6) researchers or (7) students.)ST: I question whether many, or even most, of these alleged benefits of IRs really are such. Why is an IR needed for any individual performance assessment?
The main purpose of depositing refereed journal articles is (8) to maximize their accessibility, so they are accessible to all would-be users, not just those whose institutions happen to have a subscription to the journal in which they were published. That way (9) the usage and impact of the institutional research output is maximized (and so is (10) overall research progress).
A faculty member simply submits relevant publications for the P&T committee to review; having them on an IR doesn't seem much of a benefit here.The IR provides them in digital form, in a standard way, and along with rich performance metrics: http://trac.eprints.org/projects/irstats http://citebase.eprints.org/
They don't *need* to, but it sure makes it easier, especially as CVs are increasingly online and standardized. It also helps to have the links to the full texts and metrics.And why would anyone need to go to an IR to generate a CV?
And window shopping? I submit that faculty, grad students, and researchers will go look for the work of the specialists in the areas they are most interested in,Where? How? And how do they know what work by whom at the institution to seek, where? That's precisely what IR window-shopping is for.
not canvass a wide swathe of an institution's publications, and then rely on more or less "objective" indicators of prestige ranking of departments by various bodies that conduct regular assessments.No? Even in today's world of Google ranking? I think your views may be a tad phase-lagged, Sandy.
Who is going to try using an IR to measure "the usage and impact of the institutional research output" overall?Well, the UK's Research Assessment Exercise, now converted to metrics, and the Australian RQF are examples, and this will only grow.
As we all know, usage statistics are only one small part of an assessment exercise, and they do not even exist for large parts of a university's output outside the sciences.Usage stats in particular, and metrics in general, are a part, but a growing part, of assessment, in all fields. Book metrics are on the way for book-based fields. As data-archiving grows, data-usage metrics will be developed too. None of this will (or should) go away. IRs and OA are the natural complements and conduits for this.
How does an IR help measure the success of an arts and architecture school, for instance?Patience. We're talking about peer-reviewed research in the first instance. Insofar as any discipline publishes in refereed journals at all, the very same principle applies to them. Books are next. And metrics for other forms of research output will be developed soon enough.
Given the widely disparate nature of materials that would be contained in any university-wide IR, I can't see how anyone could readily come up with an overall measure of a university's contribution to research and scholarship. This is pie-in-the-sky thinking, in my opinion.Patience. For the time being, settle for the existing and evolving metrics of refereed journal article output. The rest will come, but is not yet at issue, so there is no point invoking it by way of trying to stave off the former.
...the peer-reviewed versions of articles you are talking about being deposited have not undergone any copyediting, and my guess is that most of what gets posted will be versions that have not had copyediting done on them. I keep beating this drum, but I need to remind people that unedited faculty prose is often not something they would want to have exposed to the wider world. You mentioned that high energy physics hasn't seem to have suffered any from its exposure on arXiv, but then how many people beyond specialists actually read anything on this site?(1) We are talking about refereed journal articles, not books.
Harvard, on the other hand, is mandating deposit of the writings of its faculty in the humanities and social sciences, which at least in principle could be of interest and accessible to a nonspecialist public. My contention is that Harvard will more likely be embarrassed by the unedited writing of some of its faculty than gain anything in prestige from it.We'll see whether there is any embarrassment at Harvard. (We've had no embarrassment at Southampton ECS 'lo these half dozen years since we adopted the first of the planet's OA self-archiving mandates.) But let me correct the misapprehension that has partly been created by the ATA's otherwise commendable and successful lobbying on behalf of the NIH mandate: The rationale and urgency of OA is *not* a desperate need of access to refereed research on the part of the lay tax-paying public (other than in health-related fields). It is peer to peer access for the sake of research progress and applications for the benefit of the lay tax-paying public (in all fields).
Indeed, I can see some bloggers starting to award annual prizes to the worst writing on Harvard's site, along the lines of the Congressional "golden fleece" awards or the "razzies" that are given out each year to the worst movies.Sounds like fun. The golden fleece was mostly a fleece, but occasionally it did manage to expose research nonsense, and in that case, more power to it.
Really, believe me when I say, Stevan, that excellent scholars are not always, or even often, the best writers; many a reputation has been saved by the good work of unheralded copyeditors working behind the scenes. To protect its reputation, Harvard might find itself having to hire copyeditors itself to spruce up the articles before they get posted, or at least proofreaders who could remove the most egregious errors and typos.I really believe this when it comes to book drafts, Sandy, but book drafts are not the target of OA or OA IRs, or Green OA Self-Archiving Mandates. The target is peer-reviewed, accepted research journal articles.